IL-PCSR: Legal Corpus for Prior Case and Statute Retrieval

ACL ARR 2025 February Submission6111 Authors

Abstract: Identifying relevant statutes and prior cases/precedents for a given legal case are two of the most common tasks exercised by legal practitioners. Researchers till date have addressed the two tasks independently, thus developing completely different datasets and models for each of the task, making it difficult to compare models across both tasks despite both being legal document retrieval problems. Given the paucity of such corpora, in this resource paper, we propose a new corpus IL-PCSR (Indian Legal corpus for Prior Case and Statute Retrieval), which is a unique corpus that provides a common testbed for developing models for both the tasks (Statue Retrieval and Precedent Retrieval). We experiment extensively with several baseline models on the tasks, including lexical models and semantic models. Results show that the ensemble of a semantic model (GNN) and a lexical model (BM25) gives the best performance.

Paper Type: Long

Research Area: Resources and Evaluation

Research Area Keywords: Legal NLP, Prior Case Retrieval, Legal Statute Retrieval

Contribution Types: NLP engineering experiment, Data resources, Data analysis

Languages Studied: English

Submission Number: 6111